> Oh by the way... on installation procedures. Once I have the base file
> systems created, the next step would be to install the distribution
> kits. I am familiar with NetBSD on an Intel platform. There, what I 
> had was a shitload of files named BASE.00 thru BASE.XX, and I would
> just use Extract, and mindless feed in floppies. It seems that there is
> an Extract command in this version, but I'll be damned if I could figure
> it out. Basically all I did was:
> 
> tar --extract --directory / --verbose --gunzip /mnt/base.tgz
> 
> (from memory, I may have the parameters mixed up. You get the idea,
> I am untarring (and gunzipping) the file on my GEM partition (mounted
> on /mnt), to my root directory)
>From the INSTALL document:
    Next goto the directory in which you stored the
      distribution sets.  If e.g. you stored them in the
      root directory of the partition:

         cd /mnt/gemdos

      When there, run "Set_tmp_dir" and choose the default
      temporary directory, by hitting return at the
      prompt.

      Run the "Extract" command, giving it as its sole
      argument the name of the distribution set you wish
      to extract.  For example, to extract the base
etc, etc..
This is the way I did it myself ;-) In the newest INSTALL doc & procedures,
the Extract command asks you if you are installing an update-set. They only
thing that might be unclear in the quoted text is, that you should enter
the directory name where the sets are stored as answer to the Set_tmp_dir.
In the given example, this works out ok though.

> 
> Regarding the mail you just sent out:
> 
> > 
> >     - Atari specific docs/manual pages.
> > 	Some of the topics:
> > 	    - The mouse device
> > 	    - Using disklabels
> > 	    - Keyboard mapping
> > 	    - boot process
> > 	    - (virtual) console
> > 	    - floppy
> 
> I could offer my services on these. I've written plenty of docs for
> both personal and professional use, and consider myself an adequate
> tech writer...
Offer accepted ;-) Most, if not all, docs should be formatted as manual
pages. I think you need some input from me too on some subjects. I'll
send you some mail about it.

> 
> >     - X
> > 	I have had some private mail about this. The current status is:
> > 	    X _is_ running on NetBSD/Atari but in mono only. I have
> > 	    only tested it on ST-HIGH, but I think TT-HIGH should also
> > 	    work. I have hacked the Xdaniver server from NetBSD/Amiga
> > 	    a little bit to accomplish this. Currently Adventurer/Explorer/...
> > 	    is the only one running it besides me. I can post some instructions
> > 	    to the list how to set it up. It really should be made to run
> > 	    in TT-MED and/or a Falcon mode like that, to make it usable for
> > 	    people unable to use TT-HIGH. I don't have enough time to spare
> > 	    to do development on X currently, as I feel that I should
> > 	    better put my effort in the NetBSD/Atari 'core'.
> 
> I would like to try out X. You know (or may) that I am running my Falcon
> at ST-High. I also have access to a TT, but can only run that at 
> TT-Med, as I don't have a display capable of TT-High (it's also got a
> Nova board in it, but since I have only a standard VGA monitor, it's
> not being used).
It's not only a matter of trying out, there is probably some digging and
programming needed to get screen depths > 1 to work.

> 
> (I swap the SCSI drive between my Falcon and TT on an as-need basis...
> gotta get myself a big SCSI drive just for use on the Falcon one of
> these days...)
> 
> >     - SCSI
> > 	There are some outstanding issues here:
> > 	  - On the Falcon, the kernel still polls for the end of DMA,
> > 	    this should be solved, because it gives too much interrupt
> > 	    latency. However, it would be nice to have a Falcon owner
> > 	    do this. Modifying a driver on a platform you don't have
> > 	    is rather tedious and frustrating, both for me and the
> > 	    one testing it. As an indication: It took 5(!) months to
> > 	    get NetBSD running on the Falcon. With most of the time spend
> > 	    to get the SCSI-driver going. I like to thank Billy, Thomas
> > 	    and Helmar for their efforts in trying all my test kernels!
> 
> I can't offer any original programming, but can give you limited 
> trouble-shooting in kernel-based stuff. Mostly this would just be simple
> and obvious fixes. However, I can compile stuff (or should be able too
> shortly!), so if you want, you can mail me a copy of the latest and
> greatest source, and I can give you my input...
You already made that offer ;-) But as I said above, this is tedious, not
at least because of the relatively long turnaround times. There are no
quick tryouts possible in this sceme. Someone with a Falcon and some
programming knowlege could probably fix this in about a day or 2 while
I would need a few weeks.... The code is already present, I probably made
a thinko somewhere.

Leo.